There has been a growing and notable tendency of late amongst those who declare themselves open to tax reform to benefit developing countries to, at the same time, raise considerable doubts about available estimates of both the losses from tax abuse resulting to those countries and the gains to them that might arise from reform, which are not, of course, the same.

I have prepared no offshore estimates for a number of years now, but did at one time. And I know quite a lot about the problems involved in estimating tax losses within economies. So let me explain some of the issues.

First, any estimate of tax losses is always wrong if you mean by 'right' that it is accurate with a high degree of precision. Measuring losses always involved measuring the shadows that illicit activities cast on the economy and this will never be a precise science. Some seem quite unaware of this.

Second, this activity involves macro i.e. total economy data and yet time and again I see evidence that really rather clever accountants and micro-economists have no clue of how things work, let alone are measured, at this level where the first requirement of analysis is to understand that a great deal of what goes on appears to be the polar opposite of micro understanding. That's not chance: at a macro level the impact of transactions is often the effective double entry of what happens at the micro level.

Third, there have been few attempts at these exercises: working out what to do is hard when others have not done what you're seeking to achieve. That's very little understood.

Fourth, there is a sparsity of data to deal with, and some that is available is either incredibly expensive to secure or is simply denied to NGO researchers.

Fifth, there are decidedly limited budgets in NGOs for this work and that certainly makes repetition hard.

Sixth the number with the skills and inclination to do this work are limited.

Seventh, peer review is rarely realistically available because of many of the above issues and because in the more conventional economic peer review network, even if it could be accessed, the aim is to suppress heterdox thought in support of the neo-classical hegemony that is challenged by this work and is therefore inclined to always resist it.

What does all this mean? First that thE capacity to create estimates is limited. Second that many are, admittedly, rough around the edges and are at best indications of relative scale. Third, that the capacity to create improved data clearly exists but can be hard to deliver. And fourth that critics can always, correctly, find fault without ever offering any alternative opinion.

But there is a further consequence: if the loss estimates are hard to prepare the estimates of potential recovery are even more difficult as this requires an enormous range of very broad assumptions to be made on system change that will always be subjective. That does not mean the estimates should not be prepared but that they will be even more open to interpretation.

Now let me add some important facts. First, many of these comments could be applied to a wide range of macro data from measures of GDP onwards. Don't presume the issues only relate to NGO work.

Second, the estimates prepared are not perfect (but let's be clear: none are because all estimates are imperfect, by definition) but they have very clearly indicated the existence of a real issue that is now generally accepted to exist and have suggested its scale, and whether absolutely precisely or not indications of scale gave been provided that have suggested there is an enormous

Third, the estimates could be improved but few of the critics how this could be achieved in a meaningful way or suggest how they think the resource constraint can be removed. Unless they do their commentary does not constitute constructive criticism; it is just nit picking and frankly, all too often that is the technique of the person who is trying to undermine debate without saying so, and I have no more time for that.

So, what is needed now, in my opinion?

First, I think NGOs need to say that estimates are in a range.

Second they can say they are the best available estimates, but enhancement is possible. If there are doubts they can draw attention to them (as I do with my tax gap estimates, frequently making it clear that HMRC have an alternative opinion, even if I disagree with it). This adds to, rather than subtracts from, the strength of the argument.

But much more importantly, it is time for NGOs to say they have done their service in this area and now it is time for the OECD, World Bank, IMF, UN or a widely respected more independent national agency to take this issue on and prepare the better estimates that are needed because having this data is key to managing the outcomes on the issue. Measuring something does change it. If this loss is not measured properly then the situation will persist and the lack of action will wittingly assist that persistence.

Of course preparing such estimates cost money. But so does this problem and since we now know it is real, if not yet properly quantified, the excuse for inaction no longer exists.

In other words this debate has to move on with a common, collective and coherent call for action to assess its true scale as part of the agenda for tackling the abuse. Can we agree that this is the way forward?